Mutinex has launched its Open MMM Validation Framework, an open-source, vendor-neutral toolkit for objectively testing and comparing any MMM solution.
The Framework aims to introduce transparency and accountability for those assessing the viability of a market-mixed model and sets a new bar for how marketing effectiveness is validated
Mutinex Co-Founder and CEO Henry Innis explained to Mediaweek that most market mix models have no way to prove their effectiveness
“When it comes to governance standards, it is typically associated with words like audit, and we can trace what’s happening within a given model. You don’t have that as an established standard in the market mix modelling space.
“When you don’t have governance, the standards of something can vary wildly.”
“To my mind, the most important thing here is how do we create an established governance standard? A really good example of that being done well is the large language models, which have a number of benchmarks that they use, public benchmarks for them to solve. That’s how we know that they are reliable models that do what they say.”
Innis noted that MMM looks at billions of dollars of media spend but has no equivalent governance standard globally.
“I thought it was incumbent on us to push the boat and try to create a standard.”
Innis said that Mutinex was staying ahead of critics who may call them out for ‘marking their homework’ by making the Framework open source.
“I don’t think that what we’ve done is perfect. I believe that there’ll be contributions from other people. It’s licensed under an Apache license, which means our competitors can commercialise it without referencing us.
“They can improve it if they want to. They can use the code however they want. That’s critical and crucial.”
But why make it a free-for-all for the industry?
Innis said that the Mutinex team didn’t mind it being open to the industry and added: “What we want to see is a better standard of industry. It doesn’t bother me whether our competitors credit us or not.
He noted that their competitors, including the likes of Analytic Partners and the holding companies, are doing good things to support their initiative.
“It seems like many of them do great work and are already doing a lot of these things.
“If we come together on a collaborative industry-leading standard, we’re all going to build trust together in the open. That’s going to elevate the category.
“I want the MMM category to be an enduring and lasting category. I don’t want it to be like the digital attribution category, which became a broken Wild West of everything.
“I believe by setting a standard and working with vendors and with our competitors in a way that allows us to collaborate, we will change the industry.”
Blake Rand, Group Head of Insights at Domino’s Pizza Enterprises, noted Mutinex’s transparent and collaborative approach to governance practices that ensure efficacy with the Framework.
“The development of an open-source validation solution is a clear demonstration of their commitment to transparency and their drive to advance marketing effectiveness across the industry,” he added.
Innis said: “We know it’s robust and works. That being said, there’s always going to be ways to improve this testing suite.
“We hope by making it open source, we provide that forum for collaboration rather than do it in the dark.”
‘One of the smartest people I’ve ever come across’
When asked about Mat Baxter and his new healthcare venture TMRW, Innis had nothing but praise for the former leader of Huge.
Last year, Baxter shifted from APAC CEO role at Mutinex to board advisor role less than four months into the gig.
Mat Baxter and Henry Innis
Innis said: “Matt is still a dear friend and he’s still someone I deeply respect. I think he’s going to do tremendous things in that business.
“I think the guy is probably one of the smartest people I’ve ever come across in my career and probably one of the most brilliant people I’ve come across.
“We were texting just about 20 minutes ago. He’s a good guy. I think he can do great things.”
Top image: Henry Innis